Unformatted text preview: its full scope that is, covering all the cases - and finally also to be able to present it in the analytic form,
as you will find it here. [This is the old sense of ‘analytic’, explained on page 7.]
All metaphysicians are therefore solemnly and legally suspended from their business
until they have satisfactorily answered the question: How is a priori knowledge of
synthetic propositions possible? Only an answer to this will provide them with the
credentials they must produce if we are to credit them with teaching us things in the name
of pure reason. If they can’t produce those credentials, we - as reasonable men who have
often been deceived - should flatly refuse to listen to them, without asking any more about
what they are offering.
They may want to carry on their business not as science but as an art of swaying
people with pronouncements that are good for them and agreeable to ordinary common 17
sense. They are entitled to ply this trade; but then they should speak the modest language
of rational belief, admitting that they must not claim to Ÿknow - and should not even
Ÿconjecture - anything about what lies beyond the bounds of possible experience. The
most they can legitimately do is to Ÿassume things; and even then they are not making
assumptions for theoretical purposes (for they must renounce those), but solely for
practical use, assuming whatever is needed to guide our thought and behaviour in
everyday life. That is their only chance of being useful and wise. It will be better, too, if
they give up the name ‘metaphysician’; for metaphysicians, properly so-called, aim to be
theoretical philosophers; they try to establish judgments a priori, which means necessary
judgments; so they can’t fool around with conjectures. What they assert is science or it is
nothing at all.
In now proceeding to the answer to the question ‘How is a priori knowledge of
synthetic propositions possible?, according to the analytic [old sense] method, in which we
presuppose that such knowledge through pure reason is real, we can appeal to only two
sciences, namely pure mathematics and pure natural science. Only these can represent
objects to us in intuition. If one of them should yield an item of a priori knowledge, it
could show that this knowledge is real by showing that it fits with the intuited object; and
we could then work back from the reality of this knowledge to whatever it is that makes it
In order to move on from these kinds of pure a priori knowledge, which are both
real and grounded, to the possible kind of knowledge that we are seeking, namely to
metaphysics as a science, we must take our question a little more broadly. As well as
enquiring into Ÿthe possibility of metaphysics as a science, we must also investigate Ÿthe
natural human disposition to pursue such a science. That involves looking into the a priori
thoughts that are uncritically accepted, developed, and called ‘metaphysics’. The truth of
such thoughts is under suspicion, but the thoughts themselves are natural enough; they fall
within the scope of our question because they involve the natural conditions out of which
metaphysics arises as a science.
So our main problem splits into four questions, which will be answered one by one:
1) How is pure mathematics possible?
2) How is pure natural science possible?
3) How is metaphysics possible in general?
4) How is metaphysics possible as a science?
It may be seen that the solution of these problems, though chiefly designed to present the
core of the Critique, also has an odd feature that is worth attending to separately. We are
looking to reason itself for the sources of certain sciences, doing this so that from its
performance we can assess reason’s powers as a faculty of a priori knowledge. This
procedure also brings benefit to those sciences, in respect not of their content but of their
proper use; and they throw light on the higher question about their common origin, while
also giving an occasion better to explain their own nature. 18
FIRST PART OF THE MAIN TRANSCENDENTAL PROBLEM:
How is pure mathematics possible?
Mathematics is a great and proved field of knowledge; it already has a large scope, and
there is no limit to how far it can be extended in the future; and its results are absolutely
necessary and certain, which means that they owe nothing to experience. Mathematical
propositions are pure products of reason, yet they are thoroughly synthetic. How can
human reason create such knowledge wholly a priori? Does not our mathematical faculty,
which isn’t and can’t be based on experience, presuppose some basis for a priori
knowledge? This basis must lie deeply hidden, but we might be able to discover it through
its effects - that is, through our mathematical knowledge - if we can only track down the
sources of that knowledge.
We find that all mathematical knowledge has this special feature: it must first exhibit its
concept in in...
View Full Document